Innovative techniques based on phonon-mediated devices 273the fluctuations of the excitation number, its value scales as the square root of the
energy required on the average to produce a single excitation.
Detection of non-equilibrium phonons is very attractive because it can, in
principle, provide information about interaction position (space resolution has
already been proved with this method), discrimination about different types of
interacting radiation and the direction of the primary recoil in the target material.
The last two points remain to be proved.
8.2.3 Phonon sensors
As anticipated, the commonly used phonon sensors are STs and TESs. STs consist
usually of Ge or Si small crystals with a dopant concentration slightly below the
metal–insulator transition [21, 22]. This implies a steep dependence of the sensor
resistivity on temperature at low temperatures, where the variable range hopping
conduction mechanism dominates.
TESs are much more sensitive devices, since their resistivity changes rapidly
from a finite value to zero in a very narrow temperature interval. Normally,
the superconductive film is deposited on the absorber crystal, with a typical
thickness of few hundred nanometres, and the shape is defined after deposition by
photolithography and wet etching. With a rectangular shape the normal resistance
near the critical temperature is typically between several mand several,and
SQUID technology is required for the readout, but with meander shape resistances
of∼10 kcan be obtained, and a standard voltage-sensitive preamplifier can be
used. Films are usually made of a single superconductor. (The most interesting
results have been obtained with tungsten [23, 24].) In another approach, the film
consists of two layers (a normal metal in contact with a superconductor): this
structure allows the critical temperature to be tuned.
8.3 Innovative techniques based on phonon-mediated devices
8.3.1 Basic principles of double readout detectors
An important feature of PMDs is that a high response is expected for energies
deposited by slow (<100 keV) nuclear recoils, which are difficult to detect
with conventional devices because of their scarce ionizing power. In a perfect
calorimeter, a nuclear recoil produces the same signal of a fast electron of the
same energy, since it deposits the same amount of heat. In spite of the naiveness
of this approach, it has been proven withad hocmeasurements that the detecting
efficiency for recoiling nuclei and electrons is indeed the same within 2% in
dielectric ST-PMDs [18]. In other terms, as already anticipated,Qn1for
PMDs. As a consequence, impressively low thresholds can be achieved in large
amounts of low specific heat material (typically sapphire). If properly operated
in a low radioactive background environments, these low threshold PMDs can be
very sensitive DM detectors. The CRESST experiment has installed in the Gran